JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY

VOL. 64, NO. 18, 2014

2014 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

ISSN 0735-1097/$36.00

PUBLISHED BY ELSEVIER INC.

http://dx.doi.org/10.1016/j.jacc.2014.07.985

Platelet Activation Is Associated

With Myocardial Infarction in
Patients With Pneumonia
Roberto Cangemi, MD,* Marco Casciaro, MD,* Elisabetta Rossi, MD,* Camilla Calvieri, MD,y Tommaso Bucci, MD,*
Cinzia Myriam Calabrese, MD,* Gloria Taliani, MD,z Marco Falcone, MD,x Paolo Palange, MD,x
Giuliano Bertazzoni, MD,k Alessio Farcomeni, PHD,x Stefania Grieco, MD,z Pasquale Pignatelli, MD,*
Francesco Violi, MD,* in collaboration with the SIXTUS Study Group

ABSTRACT
BACKGROUND Troponins may be elevated in patients with pneumonia, but associations with myocardial infarction (MI)
and with platelet activation are still undened.
OBJECTIVES The aim of this study was to investigate the relationship between troponin elevation and in vivo markers
of platelet activation in the early phase of hospitalization of patients affected by community-acquired pneumonia.
METHODS A total of 278 consecutive patients hospitalized for community-acquired pneumonia, who were followed up
until discharge, were included. At admission, platelet activation markers such as plasma soluble P-selectin, soluble
CD40 ligand, and serum thromboxane B2 (TxB2) were measured. Serum high-sensitivity cardiac troponin T levels and
electrocardiograms were obtained every 12 and 24 h, respectively.
RESULTS Among 144 patients with elevated high-sensitivity cardiac troponin T, 31 had signs of MI and 113 did not. Baseline
plasma levels of soluble P-selectin and soluble CD40 ligand and serum TxB2 were signicantly higher in patients who
developed signs of MI. Logistic regression analysis showed plasma soluble CD40 ligand (p < 0.001) and soluble P-selectin
(p < 0.001), serum TxB2 (p 0.030), mean platelet volume (p 0.037), Pneumonia Severity Index score (p 0.030), and
ejection fraction (p 0.001) to be independent predictors of MI. There were no signicant differences in MI rate between
the 123 patients (45%) taking aspirin (100 mg/day) and those who were not aspirin treated (12% vs. 10%; p 0.649).
Aspirin-treated patients with MIs had higher serum TxB2 compared with those without MIs (p 0.005).
CONCLUSIONS MI is an early complication of pneumonia and is associated with in vivo platelet activation and serum
TxB2 overproduction; aspirin 100 mg/day seems insufcient to inhibit thromboxane biosynthesis. (MACCE in Hospitalized
Patients With Community-acquired Pneumonia; NCT01773863) (J Am Coll Cardiol 2014;64:191725) 2014 by the
American College of Cardiology Foundation.

ommunity-acquired pneumonia (CAP) is the

SEE PAGE 1926

most common infection leading to hospitalization in intensive care units and the most

that respiratory tract infections are associated with

common cause of death associated with infectious

an

diseases (1). Epidemiological studies have shown

acute cardiovascular events (2). This link is further

increased

risk

for

the

development

From *I Clinica Medica, Sapienza University of Rome, Rome, Italy; yDepartment of Cardiovascular, Respiratory, Nephrology,
Anesthesiology and Geriatric Sciences, Sapienza University of Rome, Rome, Italy; zInfectious and Tropical Diseases Unit,
Department of Clinical Medicine, Sapienza University of Rome, Rome, Italy; xDepartment of Public Health and Infectious Diseases,
Sapienza University of Rome, Rome, Italy; and the kUOC Emergency Medicine, Department of Internal Medicine and Medical
Specialties, Sapienza University of Rome, Rome, Italy. See the Appendix for a list of the members of the SIXTUS Study Group. This
work was supported by a grant from Sapienza University of Rome (Progetto Universitario 2010) to Prof. Violi. The authors have
reported that they have no relationships relevant to the contents of this paper to disclose.
Listen to this manuscripts audio summary by JACC Editor-in-Chief Dr. Valentin Fuster.
You can also listen to this issues audio summary by JACC Editor-in-Chief Dr. Valentin Fuster.
Manuscript received April 18, 2014; revised manuscript received June 17, 2014, accepted July 8, 2014.

of

1918

Cangemi et al.

JACC VOL. 64, NO. 18, 2014

NOVEMBER 4, 2014:191725

Platelet Activation and MI in Pneumonia

ABBREVIATIONS

supported by studies indicating that inu-

fullled the following criteria in the study: 1) age $ 18

AND ACRONYMS

enza vaccination is associated with reduced

years; 2) clinical presentation of an acute illness with

risk for hospitalization for pneumonia, heart

one or more of the following signs or symptoms sug-

disease, cerebrovascular disease, and the

gesting CAP: presence of rales, rhonchi, bronchial

risk for death from all causes during inuenza

breath sounds, dullness, increased fremitus and ego-

seasons in elderly patients (3).

phony, fever (>38.0 C), tachycardia, chills, dyspnea,

ASA = acetylsalicylic acid

CAP = community-acquired
pneumonia

CHD = coronary heart disease

A prospective study (4) focused attention

coughing (productive or unproductive cough), and

on the association of pneumonia with ele-

chest pain; and 3) presence of new consolidation(s)

vations of cardiac enzymes (e.g., creatine

on chest x-ray. Pneumonia was considered CAP if it

cardiac troponin T

kinase-MB) and signs of myocardial ischemia,

was diagnosed on hospitalization and the patient had

IQR = interquartile range

suggesting that myocardial infarction (MI)

not been discharged from an acute care facility

may be a complication of pneumonia in the

within 14 days preceding the clinical presentation.

COPD = chronic obstructive

pulmonary disease

hs-cTnT = high-sensitivity

MI = myocardial infarction

early phase of clinical presentation. More

Patients were excluded from the study if any of the

recently, Chang et al. (5) showed elevated

following criteria applied: radiographic evidence of

serum troponins (a more sensitive marker of

preexisting inltrates, severe sepsis or immunosup-

cardiac damage than creatine kinase-MB) in

pression (human immunodeciency virus infection,

Index

patients with pneumonia, but the relation-

chemotherapy, high doses of immunosuppressive

sCD40L = soluble CD40 ligand

ship with MI was not investigated. Further-

agents such as prednisone), presence of malignancy,

sP-selectin = soluble

more, the mechanism of MI in patients with

pregnancy or breastfeeding, documented severe

p-selectin

pneumonia is still elusive.

allergy to antibiotics, and health careassociated

MPV = mean platelet volume

PAD = peripheral arterial
disease

PSI = Pneumonia Severity

T2DM = type 2 diabetes

Platelets play a key role in the occurrence

mellitus

of MI, as shown by interventional trials

TxB2 = thromboxane B2

showing

that

acetylsalicylic

acid

(ASA),

which inhibits platelet thromboxane A2 production

via irreversible acetylation of COX-1 (6), lowers cardiovascular events in patients with acute or chronic
coronary heart disease (CHD) (7). In patients with
pneumonia or other infections, previous studies
showed in vivo elevation of platelet activation biomarkers, suggesting potential interplay between
platelet overactivation and cardiovascular events
during pneumonia. Thus, we speculated that in
pneumonia, differences in platelet activation might
be detected in patients with and without signs of MI.
For this purpose, we performed a prospective study in
which the relationships between in vivo markers of
platelet activation were analyzed in the early phase of
hospitalization of patients affected by CAP. Markers
analyzed included plasma levels of soluble CD40
ligand (sCD40L) (8) and soluble P-selectin (sP-selectin) (9); serum levels of thromboxane B 2 (TxB2 ), which
reects maximal platelet formation of thromboxane
A2 (1012); and serum levels of high-sensitivity cardiac troponin T (hs-cTnT).

METHODS
PATIENT SELECTION. The study was conducted at

4 centers of the University Hospital Policlinico

Umberto I (Rome, Italy). All patients with CAP admitted
to the 4 units through the emergency department from
October 2011 to April 2013 were prospectively recruited

pneumonia (13).
BASELINE ASSESSMENT. Data on demographic charac-

teristics and comorbidities were collected. Severity of

illness at presentation was quantied by the Pneumonia Severity Index (PSI), a validated prediction
score for 30-day mortality in patients with CAP
(14). Immediately after diagnosis of CAP, routine
blood laboratory tests including platelet count and
mean platelet volume (MPV), serum hs-cTnT and
high-sensitivity C-reactive protein, serum TxB2 , and
arterial blood gas test, were performed. Thereafter, serum hs-cTnT was assessed every 12 h, and
12-lead electrocardiography was repeated every 24 h.
M-mode and 2-dimensional color Doppler echocardiography was performed within 2 days of hospital
admission. Ejection fraction was measured using the
modied Simpsons rule.
Type 2 diabetes mellitus (T2DM), hypertension,
history of CHD, dyslipidemia, peripheral arterial disease (PAD), and chronic obstructive pulmonary disease (COPD) were dened as previously described
(1517). Baseline treatments were dened according
to patients pharmacological histories. Patients already treated with ASA before admission were categorized as ASA users. Compliance with ASA and other
medications was monitored daily.
This study was conducted according to the principles stated in the Declaration of Helsinki. The
institutional review board approved this prospective, observational study, which was registered at
ClinicalTrials.gov (NCT01773863).

and followed up. After they gave written informed

STUDY ENDPOINTS. The primary study end point

consent, we enrolled 278 consecutive patients who

was the occurrence of MI during the hospital stay.

Cangemi et al.

JACC VOL. 64, NO. 18, 2014

NOVEMBER 4, 2014:191725

Platelet Activation and MI in Pneumonia

MI criteria were those of the third universal de-

and a coefcient of variation of <10% is achieved

nition of MI: the detection of an increase in cardiac

at 0.013 m g/l (19).

troponin with at least 1 value above the 99th per-

STATISTICAL ANALYSIS. Categorical variables are

centile upper reference limit, associated with at least

reported as counts and percentages and continuous

1 of the following: 1) chest pain; 2) detection of new or

variables as mean  SD or medians and interquartile

presumably

ST-segmentT-wave

ranges (IQRs). Differences between percentages were

changes or new left bundle branch block; 3) devel-

assessed by chi-square or Fisher exact tests. Student

opment of pathological Q waves on electrocar-

unpaired t tests and analysis of variance were used for

diography; 4) de novo imaging evidence of viable

normally distributed continuous variables. Appro-

myocardium loss or regional wall motion abnormal-

priate nonparametric tests (Mann-Whitney, Kruskal-

ity; 5) identication of an intracoronary thrombus by

Wallis, and Spearman rank correlation tests) were

angiography or autopsy; and 6) cardiac death with

used for all other variables.

new

signicant

symptoms suggestive of myocardial ischemia and

To better dene the relationships between TxB 2,

presumed new ischemic electrocardiographic changes

sCD40L, and sP-selectin levels and cardiovascular

or new left bundle branch block (18).

events, we grouped these continuous marker mea-

ST-segment elevation MI and nonST-segment

elevation MI were dened as previously reported (18).
Therapeutic treatment at admission and during the
hospital stay were registered.

surements into 4 categories, with cutoffs on the basis

of quartiles of the observed measurements.
The bivariate and multivariate effects of prognostic factors and treatments on the primary and

Adjudication of cardiovascular events was per-

secondary endpoints were assessed by means of

formed by a committee of cardiologists (C.C. and A.C.)

logistic regression models. Wald condence intervals

who did not participate in patient recruitment and

and tests for odds ratios and adjusted odds ratios

follow-up and were unaware of any patients clinical

were computed on the basis of the estimated stan-

and laboratory characteristics.

dard errors. In addition to TxB2, sCD40L, and

PLASMA LEVELS OF sCD40L, sP-SELECTIN, AND

SERUM TXB 2 . Blood samples to measure biomarkers

of platelet activation were obtained within 24 h of

admission and at the end of hospitalization. Samples
with or without 3.8% sodium citrate were taken
without stasis from patients who had fasted for at
least 12 h.
After withdrawal of a rst blood sample, 1-ml
serum aliquots were immediately transferred into
glass tubes and allowed to clot at 37  C for 1 h. Serum
was separated by centrifugation and frozen at 80  C
until assayed. Serum TxB 2 was measured using an
enzyme-linked immunosorbent assay commercial kit
(R&D Systems, Inc., Minneapolis, Minnesota) and
expressed as nanograms per milliliter. Intra- and
interassay coefcients of variation for the TxB 2 assay
kit were 5.9% and 8.9%, respectively.
Citrated samples were centrifuged to separate
plasma, which was frozen at 80 C until assayed.
Plasma sCD40L and sP-selectin levels were measured
with a commercial immunoassay (Tema Ricerca,
Castenaso, Italy). Intra- and interassay coefcients of
variation were 5% and 7% for sCD40L and 4.3% and
6.1% for sP-selectin.
SERUM

HS-cTNT

MEASUREMENT. Hs-cTnT

sP-selectin levels, possible independent variables

considered were age, sex, body mass index, PSI
score, ejection fraction, history of CHD or stroke,
T2DM, hypertension, renal failure, COPD, chronic or
paroxysmal atrial brillation, dyslipidemia, PAD,
high-sensitivity C-reactive protein, and use of statins
and ASA.
For multivariate models, model selection was
performed using forward stepwise regression on the
basis of the Akaike information criterion.
Variance ination factors were computed to assess
collinearity for each predictor in the nal multivariate
model, with a variance ination factor < 2.5 considered safe. The variance ination factor criterion was
satised in all reported models, hence there were no
collinearity issues.
TxB 2, sCD40L, and sP-selectin levels evaluated at
the end of hospitalization were compared with baseline levels by Wilcoxon rank sum tests.
Only p values <0.05 were regarded as statistically
signicant. All tests were 2-tailed, and analyses were
performed using computer software packages (R
version 2.15.2, R Development Core Team, Vienna,
Austria).
S a m p l e s i z e . The minimum sample size was com-

levels

puted considering 1) an expected rate of MI of 5%

were measured with the Elecsys 2010 (Roche Di-

(2024); 2) a relevant difference in TxB2 levels to be

agnostics, Indianapolis, Indiana) at a dedicated core

detected between groups of jd j $ 50 ng/ml; 3) a SD

laboratory. According to the manufacturer, the

between the groups of 35 ng/ml; and 4) type I error

99th-percentile cutoff point for hs-cTnT is 0.014 m g/l,

probability of a 0.05 and power of 1  b 0.90.

1919

Cangemi et al.

1920

JACC VOL. 64, NO. 18, 2014

NOVEMBER 4, 2014:191725

Platelet Activation and MI in Pneumonia

RESULTS
CAP patients (n=278)

We recruited 278 patients hospitalized for CAP (173

men, 105 women; mean age 70.0  15.7 years).
Patients with
hs-cTnT0.014 (n=134)

Most of the patients had arterial hypertension

Patients with
hs-cTnT>0.014 (n=144)

(68%). Histories of CHD were present in 36% of

patients, previous strokes in 12%, T2DM in 24%,
COPD in 35%, PAD in 7%, and dyslipidemia in 21%.

Patients without signs of MI

(n=113)

Patients with signs of MI

(n=31)

Histories of paroxysmal atrial brillation were present in 12% of patients, while 16% were affected by
chronic (persistent or permanent) atrial brillation,

F I G U R E 1 Study Flow Diagram

and 18% had severe renal failure (i.e., glomerular

CAP community-acquired pneumonia; hs-cTnT high-

ltration rate <30 ml/min). Among the entire pop-

sensitivity cardiac troponin T; MI myocardial infarction.

ulation, 123 patients (47%) were treated with ASA

100 mg/day.
Elevated serum levels of hs-cTnT (>0.014 m g/l;

This resulted in a sample size of 220 patients. The

median 0.042 m g/l; IQR: 0.024 to 0.116 mg/l) were

expected event rate and TxB2 levels were obtained

found in 144 of the patients with CAP (52%). Of these

by analyzing data from previous studies (25).

144 patients, 113 had isolated hs-cTnT elevations, and

31 showed signs of MI (Figure 1). MI generally
occurred within 48 h of pneumonia clinical presen-

T A B L E 1 Patient Characteristics

Variable

Age, yrs

tation. Most MIs were nonST-segment elevation MIs

(n 26), 2 were ST-segment elevation MIs, 3 were

CAP With hs-cTnT

#0.014 mg/l
(n 134)

CAP With hs-cTnT

>0.014 m g/l
Without MI
(n 113)

CAP With hs-cTnT

>0.014 mg/l
and MI
(n 31)

p Value

61.8  16.6

77.3  10.2

79.2  8.9

<0.001

fatal, and all but 2 had no chest pain.

Clinical

characteristics

of

patients

with

CAP

without hs-cTnT elevation (median 0.010; IQR: 0.007

62

65

58

0.742

to 0.011 m g/l), patients with hs-cTnT elevation (me-

BMI, kg/m2

26.4  4.1

26.6  3.8

26.8  3.6

0.875

Ejection fraction, %

56.4  6.2

49.7  10.4

43.9  8.6

<0.001

dian 0.033; IQR: 0.023 to 0.060 m g/l) without signs of

Men

Pre-existing comorbid
conditions
History of CHD

MI, and patients with troponin hs-cTnT elevation

(median 0.299; 0.165 to 0.556 m g/l) and signs of MI

21

45

65

<0.001

are summarized in Table 1. Patients with hs-cTnT

Previous stroke

11

23

0.104

elevation (whether or not they developed MIs) were

T2DM

16

31

32

0.015

older and had a higher prevalence of COPD, chronic

Hypertension

58

79

84

0.001

Renal failure

33

29

<0.001

COPD

atrial brillation, PAD, renal failure, and T2DM than

patients without hs-cTnT elevation.

25

48

36

0.017

PAF

16

16

0.281

CAF

25

16

0.002

Peripheral artery
disease

10

13

0.018

17

24

29

0.231

isolated hs-cTnT but without signs of MI (p < 0.001)

ASA

37

52

48

0.069

(Table 1). When ASA users and nonusers were

Statins

32

36

37

0.776

5.9 (2.712.8)

6.0 (3.312.7)

6.3 (3.511.6)

0.955

75  27

104  25

120  29

Platelet count

223 (173319)

238 (180274)

237 (173249)

MPV,

8.3 (7.58.8)

8.6 (7.79.5)

9.4 (8.210.4)

3 (2.24.0)

3.2 (2.46.0)

10 (9.513)

<0.001

15 (1119)

22 (1630)

37 (2848)

<0.001

No signicant differences in platelet count were

108 (60200)

106 (50180)

200 (100352)

0.004

observed in patients with normal or elevated hs-

Dyslipidemia

hs-CRP, ng/ml
PSI score

sCD40L, ng/ml
sP-selectin, ng/ml
TxB2, ng/ml

Baseline plasma levels of sCD40L and sP-selectin,

serum TxB 2, and MPV were signicantly higher in
patients who developed signs of MI. Furthermore,
plasma sP-selectin levels were higher in patients with

<0.001

considered separately, both sCD40L and sP-selectin

were signicantly correlated with serum TxB2 in

0.767

ASA users (Rs 0.440; p < 0.001, and Rs 0.214;

0.001

p 0.020, respectively) and nonusers (Rs 0.269;

Values are mean  SD, %, or median (interquartile range).

ASA acetylsalicylic acid; BMI body mass index; CAF chronic (persistent or permanent) atrial brillation;
CAP community-acquired pneumonia; CHD coronary heart disease; COPD chronic obstructive pulmonary
disease; hs-CRP high-sensitivity C-reactive protein; hs-cTnT high-sensitivity cardiac troponin T; MI
myocardial infarction; MPV mean platelet volume; PAD peripheral artery disease; PAF paroxysmal atrial
brillation; PSI Pneumonia Severity Index; sCD40L soluble CD40 ligand; sP-selectin soluble P-selectin;
T2DM type 2 diabetes mellitus; TxB2 thromboxane B2.

p 0.001, and Rs 0.258; p 0.001, respectively).

cTnT. Platelet count also did not correlate with

markers of platelet activation or serum TxB2 (not
shown).
To better dene the magnitude of the relationship
between markers of platelet activation and MI,
patients with CAP were stratied into sCD40L,

Cangemi et al.

JACC VOL. 64, NO. 18, 2014

NOVEMBER 4, 2014:191725

Platelet Activation and MI in Pneumonia

sP-selectin, and TxB2 quartiles (Figure 2). The largest

number of MIs occurred in the upper quartiles of
all 3 variables, corresponding to sCD40L >6.0 ng/ml,

sP-selectin >26 ng/ml, and TxB 2 >200 ng/ml.

CHD (p < 0.001), previous stroke (p 0.040), hypertension (p 0.044), sCD40L >6.0 ng/ml (p < 0.001),

25

Number of Events

In univariate analysis, age (p < 0.001), PSI score

(p < 0.001), ejection fraction (p < 0.001), history of

30

20
15
10

sP-selectin >26 ng/ml (p < 0.001), TxB 2 >200 ng/ml

(p 0.003), and MPV (p 0.001) were associated with

0
1

MI occurrence. By multivariate analysis, sCD40L,

sP-selectin, serum TxB2 , MPV, PSI score, and ejection
fraction were independently associated with MI

(Table 2).
123 (45%) were treated with ASA 100 mg/day. The
rate of MI was 10% in ASA-untreated patients and
12% in ASA-treated patients (p 0.649). Patients

20
15
10

with long-term ASA use showed lower serum TxB 2

5
0
1

ng/ml [IQR: 20 to 100 ng/ml] vs. 160 ng/ml [IQR: 106

to 247 ng/ml]; p < 0.001). Among ASA-untreated

had higher serum TxB2 values than ASA-untreated

patients (p 0.005) (Figure 3). Only 9 patients had
serum TxB2 <10 ng/ml, corresponding to full COX1
inhibition (26).

Number of Events

ASA-treated patients; ASA-treated patients with MIs

sP-Selectin Quartiles
16
14

who experienced MIs compared with those who did

not (p 0.001). Similar ndings were detected in

30

levels than non-ASA-treated patients (median 61

patients; serum TxB 2 levels were higher in those

25

Number of Events

ASA AND THROMBOXANE. Among the 278 patients,

sCD40L Quartiles

12
10
8
6
4
2
0
1

FOLLOW-UP OF PLATELET ACTIVATION BIOMARKERS.

Thromboxane B2 Quartiles

Analyses of plasma sCD40L and sP-selectin levels

and serum TxB2 were repeated at the end of hospi-

F I G U R E 2 Intrahospital MI According to

talization (within approximately 10 days of hospital

In Vivo Platelet Activation Markers

admission). Compared with baseline values, all of

the variables signicantly decreased, independent
of ASA use (Figure 4).

DISCUSSION

Number of patients with CAP with intrahospital MI stratied

according to soluble CD40 ligand (sCD40L) (A), soluble
P-selectin (sP-selectin) (B), and thromboxane B2 (C) quartiles.
Other abbreviations as in Figure 1.

In this study, we prospectively analyzed the incidence of MI in patients hospitalized for CAP. The
novel ndings of the study are as follows: 1) in the

Previous

studies

showed

that

pneumonia

is

early phase of pneumonia (within 48 h of presenta-

complicated by elevated hs-cTnT values, indicating

tion), patients showed elevated hs-cTnT, which may

an association between infectious disease and myo-

be either isolated or MI associated; 2) MI was not

cardial injury. Notably, Chang et al. (5) recently re-

associated with chest pain in the majority of cases;

ported that 19% of patients with pneumonia have

3) the severity of pneumonia, as assessed by PSI score

elevated hs-cTnT values at admission. We found that

and ejection fraction, was independently associated

elevated hs-cTnT was even more prevalent in our

with MI; 4) in vivo platelet activation and over-

population; hs-cTnT was elevated in >50%. This dif-

production of serum TxB 2 formation were indepen-

ference may be because we measured hs-cTnT levels

dently associated with MI; and 5) ASA 100 mg/day

not only at baseline but also every 12 h during the

seemed insufcient for preventing serum TxB 2

rst days of hospitalization. Increased hs-cTnT was

overproduction.

detectable up to 48 h from hospital admission,

1921

Cangemi et al.

JACC VOL. 64, NO. 18, 2014

NOVEMBER 4, 2014:191725

Platelet Activation and MI in Pneumonia

T A B L E 2 Adjusted ORs for MI According to Selected Variables

on the Basis of Logistic Regression Analysis

MI usually occurred within 48 h of clinical presentation, consistent with the study by CorralesMedina et al. (2), who found that more than 50% of

95% CI
Variable

cardiovascular complications occurred within 24 h of

OR

Lower

Upper

p Value

VIF*

sCD40L >6.0 ng/ml

1.234

1.147

1.327

<0.001

1.129

sP-selectin >26 ng/ml

1.233

1.142

1.330

<0.001

1.217

TxB2 >200 ng/ml

1.084

1.008

1.166

0.030

1.035

presentation. The peculiarity of our methodological

approach turned out to be particularly useful,
because almost all patients with pneumonia had si-

MPV

1.025

1.002

1.048

0.037

1.056

lent MIs; without daily monitoring of cardiac tropo-

Ejection fraction

0.994

0.991

0.998

0.001

1.150

nins and electrocardiography, it would not be

PSI score

1.001

1.001

1.002

0.030

1.254

possible to fully appreciate the occurrence of MI.

Among the clinical variables analyzed, age, PSI score,

After adjusting for age, sex, BMI, history of CHD, previous stroke, T2DM, hypertension, renal failure, COPD, CAF, PAF, PAD, dyslipidemia, platelet count, hs-CRP,
use of statins, and ASA. *Given that all VIFs are well below the cutoff, the parameters can be directly interpreted, and there are no collinearity issues.
CI condence interval; OR odds ratio; VIF variance ination factor; other
abbreviations as in Table 1.

ejection fraction, history of CHD, previous stroke, and

hypertension were associated with MI. However, in
multivariate analysis, only PSI and ejection fraction
were independently associated with MI. Of note, patients with a mean PSI score of 120 were at higher risk
for MI. Our observation of an association between

indicating that myocardial damage should be inves-

MI and history of CHD in 65% of patients who expe-

tigated in the early phase of hospitalization.

rienced MI is also interesting and suggests that

Among patients with elevated hs-cTnT, 31 were

diagnosed as affected by MI, yielding an 11% intra-

patients with underlying CHD are more susceptible

to MI after pneumonia.

hospital incidence of MI. Although previous studies

Few studies have analyzed platelet behavior in

showed that pneumonia may be complicated by MI (2),

pneumonia. Patients with MI complicated by pneu-

those were mostly retrospective studies, and the rate

monia displayed platelet hyperaggregation and non-

of MI ranged from as low as 0.8% to as high as 10.7%

response to ASA more than patients with MI

(2124). In the only prospective study reported thus

without pneumonia, suggesting that lung infection

far, Corrales-Medina et al. (4) found a 3.6% rate during

could trigger platelet activation (27,28). Additionally,

a follow-up period of 7 days. In the present study, we

Kreutz et al. (29) found higher platelet aggregation in

found a higher MI rate, likely because we measured

patients affected by viral upper respiratory tract in-

hs-cTnT, a more sensitive marker of myocardial ne-

fections compared with controls. The interplay be-

crosis, and because we monitored hs-cTnT values

tween platelet activation and MI in a population

daily, along with electrocardiography. Moreover, our

affected by pneumonia had never been examined.

hospitalized patients were older, had a higher preva-

We report for the rst time that in vivo platelet

lence of comorbidities, and had higher PSI scores.

activation, as assessed by plasma levels of sCD40L

and sP-selectin, was independently associated with
MI. TxB2 overproduction accompanied in vivo platelet activation, which was also independently associ-

600

Thromboxane B2 (ng/ml)

1922

500

Patients without MI
Patients with MI

ated with MI. Even if this suggests a role for

COX1 activation in enhancing platelet activation,
platelet TxB2 overproduction is unlikely to be the

400

only mechanism for platelet activation, as serum

**

300

TxB 2 was only weakly correlated with sCD40L and

sP-selectin.
Previous studies demonstrated that an increased

200

platelet count was more predictive of clinical out-

100

comes in patients with CAP than leukocyte count

abnormalities (30). However, we did not observe any

0
ASA-Untreated Patients

ASA-Treated Patients

differences in platelet count between patients with or

without MIs; conversely, MPV was both signicantly

F I G U R E 3 Thromboxane B 2 Levels in ASA-Untreated and

ASA-Treated Patients Who Did or Did Not Experience MIs

*p 0.001; **p 0.005. ASA acetylsalicylic acid;

MI myocardial infarction.

higher in patients with MIs and an independent predictor, consistent with previous studies showing that
MPV is independently associated with MI (31).
Bacteria may activate platelets; thus, the association between pneumonia and in vivo platelet

Cangemi et al.

JACC VOL. 64, NO. 18, 2014

NOVEMBER 4, 2014:191725

Platelet Activation and MI in Pneumonia

10
*

sCD40L (ng/ml)

8
6
4
2
0
ASA-Untreated Patients

ASA-Treated Patients

B
sP-selectin (ng/ml)

50
40
Cangemi, R., et al., J Am Coll Cardiol. 2014; 64(18):191725.

30
C E N T R A L I L L U S T R A T I O N Potential Mechanisms for Platelet Activation

During Pneumonia

20

Platelets can interact directly with bacteria or with lipopolysaccharide (LPS)

10

(an endotoxin) on the surface of Gram-negative bacteria by a Toll-like receptor 4

(TLR4)mediated mechanism; this results in platelet activation and aggregation and,
eventually, thrombus formation. sCD40L soluble CD40 ligand; sP-selectin soluble

0
ASA-Untreated Patients

Thromboxane B2 (ng/ml)

ASA-Treated Patients

P-selectin.

450
*

400

receptors (28). However, whether either a specic

350

agent or the infection burden is implicated in platelet

300

activation and, eventually, coronary thrombosis re-

250

mains to be explored (28). In this context, the role of

endotoxins should be examined, as there is evidence

200

to support a role for endotoxins as platelet stimula-

150

tors

100

via

Toll-like

receptor

(32,33)

(Central

Illustration). An intriguing nding of the present

50

study was that although patients on ASA treatment

0
ASA-Untreated Patients
Hospital admission

ASA-Treated Patients
End of hospitalization

showed lower serum TxB2 compared with nonusers,

serum TxB 2 in ASA-treated patients was persistently
elevated, suggesting that low-dose ASA was insufcient to fully inhibit COX1. Accordingly, many pa-

F I G U R E 4 In Vivo Platelet Activation Markers at

tients on ASA had values of serum TxB2 >10 ng/ml,

Hospital Admission and at the End of Hospitalization

while optimal long-term ASA treatment is usually

Plasma levels of soluble CD40 ligand (sCD40L) (A), soluble

associated with serum TxB2 <10 ng/ml (26). Further-

P-selectin (sP-selectin) (B), and serum thromboxane B2 (C)

more, ASA-treated patients who experienced MIs had

at hospital admission and at the end of hospitalization in acetylsa-

signicantly higher serum TxB 2 than ASA-treated

licylic acid (ASA)untreated and ASA-treated patients. *p <0.001.

patients who did not experience MIs.

Among the whole study population, 113 patients
(41%) displayed elevated hs-cTnT not associated with

activation is biologically plausible. In particular,

electrocardiographic changes, which were considered

bacteria-platelet interaction may occur via bacteria

isolated hs-cTnT elevation. Isolated elevation of hs-

binding to platelets, either directly, through a bacte-

cTnT can be detected in several clinical settings not

rial surface protein, or indirectly by a plasma-bridging

associated with MI, including infections and sepsis,

molecule

and is usually considered a specic marker of

linking

bacterial

and

platelet

surface

1923

1924

Cangemi et al.

JACC VOL. 64, NO. 18, 2014

NOVEMBER 4, 2014:191725

Platelet Activation and MI in Pneumonia

myocardial injury (34). Even if elevated hs-cTnT is a

a dose of 100 mg ASA twice daily is more effective

marker of poor prognosis (35), its use for diagnosis

than ASA 100 mg/day to fully prevent platelet TxB 2

and management in settings unrelated to acute cor-

(36). We cannot exclude that a similar phenomenon

onary syndrome is still evolving (34). Of note, with

may also occur in pneumonia and that another ASA

the exception of a weak correlation with sP-selectin,

regimen should therefore be adopted in this clinical

neither sCD40L nor serum TxB2 was correlated with

setting. However, we cannot exclude that the

isolated hs-cTnT.

persistent platelet TxB 2 elevation, despite ASA treat-

STUDY LIMITATIONS. The study had limitations and

ment, observed at baseline, may result from lowered

implications. Because CAP affected our population,

ASA bioavailability or poor compliance (37).

the data cannot be extrapolated to other types of

pneumonia. The mechanism accounting for platelet

CONCLUSIONS

activation is not fully claried by the present study;

however, the correlation between serum TxB 2 and

We provide evidence that >50% of patients with

the 2 markers of platelet activation suggests that

pneumonia had hs-cTnT elevation, which was iso-

platelet TxB 2 overproduction might play a role. We

lated or associated with MI in about 11% of cases.

also did not investigate if pneumonia was of bacterial

Because MI was silent in the majority of cases, daily

or viral origin or if specic pathogens or systemic

monitoring of troponins and electrocardiography is

inammation were responsible for platelet activation

needed for detection. Platelet activation was signi-

and platelet TxB 2 overproduction. However, acute

cantly associated with MI, suggesting a role for

infection per se is likely to be implicated, as in vivo

platelets in triggering coronary thrombosis. Daily 100-

platelet activation and serum TxB2 signicantly de-

mg doses of ASA were unable to prevent platelet TxB 2

creased at hospital discharge, independent of ASA

formation. Future studies should be done to identify

use. Finally, we have no explanation for MI being

a more appropriate ASA dose to fully block COX1,

silent in patients with pneumonia, and further study

thus preventing serum TxB 2 overproduction in pa-

should be done to clarify this phenomenon. We

tients with pneumonia.

cannot exclude that an imbalance between myocardial oxygen supply and demand was the primary

REPRINT REQUESTS AND CORRESPONDENCE: Prof.

cause of MI in many of our critically ill patients

Francesco Violi, I Clinica Medica, Sapienza University

(i.e., MI type 2 according to the universal denition

of Rome, Viale del Policlinico 155, Roma 00161, Italy.

[18]). Therefore, further study in an even larger

E-mail: francesco.violi@uniroma1.it.

population is needed to better appreciate the rate of

MI in this setting.
An important implication of this study is that in
patients with severe pneumonia, such as those with
PSI scores of approximately 120, corresponding to
classes IV and V (14), or prior CHD, daily monitoring
of cardiac troponins and electrocardiography is
necessary to detect MI, as MI was rarely associated
with chest pain. Such monitoring should be done
immediately

after

pneumonia

diagnosis,

as

MI

generally occurred within 48 h of presentation.

Another important issue raised by this study is the
inefcacy of ASA 100 mg/day in completely preventing serum TxB2 formation in patients with pneumonia, suggesting that this dosage does not fully
inhibit COX-1. In other clinical settings, it is becoming

PERSPECTIVES
COMPETENCY IN MEDICAL KNOWLEDGE:
Elderly patients are at risk for developing myocardial
ischemia and infarction early in the course of CAP.
Elevated platelet activation markers and thromboxane
A2 generation in these patients suggest that platelets
play a role in the pathogenesis of these cardiac
events.
TRANSLATIONAL OUTLOOK: Future studies
should address the safety and efcacy of administering ASA or other antiplatelet drugs to patients
presenting with acute CAP.

evident that because of accelerated platelet turnover,

REFERENCES
1. McCabe C, Kirchner C, Zhang H, et al. Guidelineconcordant therapy and reduced mortality and
length of stay in adults with community-acquired
pneumonia: playing by the rules. Arch Intern
Med 2009;169:152531.

2. Corrales-Medina VF, Musher DM, Shachkina S,

et al. Acute pneumonia and the cardiovascular
system. Lancet 2013;381:496505.
3. Nichol KL, Nordin J, Mullooly J, et al. Inuenza
vaccination and reduction in hospitalizations for

cardiac disease and stroke among the elderly.

N Engl J Med 2003;348:132232.
4. Corrales-Medina VF, Musher DM, Wells GA, et al.
Cardiac complications in patients with communityacquired pneumonia: incidence, timing, risk

Cangemi et al.

JACC VOL. 64, NO. 18, 2014

NOVEMBER 4, 2014:191725

factors, and association with short-term mortality.

Circulation 2012;125:77381.
5. Chang CL, Mills GD, Karalus NC, et al. Biomarkers of cardiac dysfunction and mortality from
community-acquired pneumonia in adults. PLoS
One 2013;8:e62612.
6. FitzGerald GA. Mechanisms of platelet activation: thromboxane A2 as an amplifying signal
for other agonists. Am J Cardiol 1991;68 Suppl:
11B5B.
7. Antithrombotic Trialists Collaboration. Collaborative meta-analysis of randomised trials of
antiplatelet therapy for prevention of death,
myocardial infarction, and stroke in high risk
patients. BMJ 2002;324:7186.
8. Angelico F, Alessandri C, Ferro D, et al.
Enhanced soluble CD40L in patients with the
metabolic syndrome: relationship with in vivo
thrombin generation. Diabetologia 2006;49:
116974.
9. Blann AD, Nadar SK, Lip GY. The adhesion
molecule P-selectin and cardiovascular disease.
Eur Heart J 2003;24:216679.
10. Patrignani P, Filabozzi P, Patrono C. Selective
cumulative inhibition of platelet thromboxane
production by low-dose aspirin in healthy subjects. J Clin Invest 1982;69:136672.

Platelet Activation and MI in Pneumonia

16. Loffredo L, Marcoccia A, Pignatelli P, et al.

Oxidative-stress-mediated arterial dysfunction in
patients with peripheral arterial disease. Eur Heart
J 2007;28:60812.

reactivity and activation: an explanation for the

higher rates of myocardial infarction and stroke
during viral illness. J Thromb Haemost 2005;3:
21089.

17. Rabe KF, Hurd S, Anzueto A, et al., Global

Initiative for Chronic Obstructive Lung Disease.

30. Mirsaeidi M, Peyrani P, Aliberti S, et al.

Thrombocytopenia and thrombocytosis at time of

Global strategy for the diagnosis, management,

and prevention of chronic obstructive pulmonary
disease: GOLD executive summary. Am J Respir
Crit Care Med 2007;176:53255.

hospitalization predict mortality in patients with

community-acquired pneumonia. Chest 2010;137:
41620.

18. Thygesen K, Alpert JS, Jaffe AS, et al. Third

universal denition of myocardial infarction. J Am
Coll Cardiol 2012;60:158198.
19. Reichlin T, Hochholzer W, Bassetti S, et al.
Early diagnosis of myocardial infarction with sensitive cardiac troponin assays. N Engl J Med 2009;
361:85867.
20. Viasus D, Garcia-Vidal C, Manresa F, et al. Risk
stratication and prognosis of acute cardiac events
in hospitalized adults with community-acquired
pneumonia. J Infect 2013;66:2733.
21. Ramirez J, Aliberti S, Mirsaeidi M, et al. Acute
myocardial infarction in hospitalized patients with
community-acquired pneumonia. Clin Infect Dis
2008;47:1827.
22. Musher DM, Rueda AM, Kaka AS, et al. The
association between pneumococcal pneumonia

11. Capone ML, Tacconelli S, Sciulli MG, et al.

and acute cardiac events. Clin Infect Dis 2007;45:

15865.

Clinical pharmacology of platelet, monocyte, and

vascular cyclooxygenase inhibition by naproxen
and low-dose aspirin in healthy subjects. Circulation 2004;109:146871.

23. Corrales-Medina VF, Serpa J, Rueda AM, et al.

Acute bacterial pneumonia is associated with the
occurrence of acute coronary syndromes. Medicine
(Baltimore) 2009;88:1549.

12. De Caterina R, Giannessi D, Bernini W, et al.

24. Mandal P, Chalmers JD, Choudhury G, et al.

Vascular complications are associated with poor
outcome in community-acquired pneumonia. QJM
2011;104:48995.

Low-dose aspirin in patients recovering from

myocardial infarction. Evidence for a selective inhibition of thromboxane-related platelet function.
Eur Heart J 1985;6:40917.
13. Venditti M, Falcone M, Corrao S, et al., Study
Group of the Italian Society of Internal Medicine. Outcomes of patients hospitalized with
community-acquired, health care-associated,
and hospital-acquired pneumonia. Ann Intern
Med 2009;150:1926.
14. Aujesky D, Fine MJ. The pneumonia severity
index: a decade after the initial derivation and
validation. Clin Infect Dis 2008;47(suppl):
S1339.
15. Perk J, De Backer G, Gohlke H, et al., European
Association for Cardiovascular Prevention &
Rehabilitation (EACPR); ESC Committee for Practice Guidelines (CPG). European guidelines on
cardiovascular disease prevention in clinical practice (version 2012). The Fifth Joint Task Force of
the European Society of Cardiology and Other
Societies on Cardiovascular Disease Prevention in
Clinical Practice (constituted by representatives of
nine societies and by invited experts). Eur Heart J
2012;33:1635701.

25. Pignatelli P, Di Santo S, Barilla F, et al. Multiple anti-atherosclerotic treatments impair aspirin
compliance: effects on aspirin resistance.
J Thromb Haemost 2008;6:18324.
26. Frelinger AL III, Furman MI, Linden MD,
et al. Residual arachidonic acid-induced platelet
activation via an adenosine diphosphatedependent but cyclooxygenase-1- and cyclooxygenase-2-independent pathway: a 700-patient
study of aspirin resistance. Circulation 2006;113:
288896.
27. Modica A, Karlsson F, Mooe T. Platelet aggregation and aspirin non-responsiveness increase when an acute coronary syndrome is
complicated by an infection. J Thromb Haemost
2007;5:50711.
28. Fitzgerald JR, Foster TJ, Cox D. The interaction
of bacterial pathogens with platelets. Nat Rev
Microbiol 2006;4:44557.
29. Kreutz RP, Bliden KP, Tantry US, et al. Viral
respiratory tract infections increase platelet

31. Martin JF, Kristensen SD, Mathur A, et al. The

causal role of megakaryocyte-platelet hyperactivity in acute coronary syndromes. Nat Rev Cardiol 2012;9:65870.
32. Berthet J, Damien P, Hamzeh-Cognasse H,
et al. Human platelets can discriminate between
various bacterial LPS isoforms via TLR4 signaling
and differential cytokine secretion. Clin Immunol
2012;145:189200.
33. Stark RJ, Aghakasiri N, Rumbaut RE. Plateletderived Toll-like receptor 4 (Tlr-4) is sufcient to
promote microvascular thrombosis in endotoxemia. PLoS One 2012;7:e41254.
34. de Lemos JA. Increasingly sensitive assays
for cardiac troponins: a review. JAMA 2013;309:
22629.
35. Saunders JT, Nambi V, de Lemos JA, et al.
Cardiac troponin T measured by a highly sensitive assay predicts coronary heart disease, heart
failure, and mortality in the Atherosclerosis Risk
in Communities Study. Circulation 2011;123:
136776.
36. Capodanno D, Patel A, Dharmashankar K, et al.
Pharmacodynamic effects of different aspirin
dosing regimens in type 2 diabetes mellitus patients with coronary artery disease. Circ Cardiovasc
Interv 2011;4:1807.
37. Grosser T, Fries S, Lawson JA, et al. Drug
resistance and pseudoresistance: an unintended
consequence of enteric coating aspirin. Circulation
2013;127:37785.

KEY WORDS cardiovascular disease,

human, platelet aggregation, prospective
studies, risk factors

A PP END IX The members of the SIXTUS

Study Group are: Fabiana Albanese, MD, Elisa
Biliotti, MD, Roberto Carnevale, PHD, Elisa
Catasca, MD, Andrea Celestini, MD, Rozenn
Esvan, MD, Lucia Fazi, MD, Paolo Marinelli, MD,
Michela Mordenti, MD, Laura Napoleone, MD,
Michela Palumbo, MD, Daniele Pastori, MD,
Ludovica Perri, MD, Marco Proietti, MD, Rivano
Capparuccia Marco, MD, Alessandro Russo, MD,
Roberta Russo, MD, Valentino Sarallo, MD,
Gabriele Salvatori, MD, Maria Gabriella
Scarpellini, MD, and Ines Ullo, MD.

1925